Merging power and data: good, bad, or both?

-April 25, 2017

Putting power on a copper cable originally or primarily designed for signals is not a new concept, nor is the reverse situation. Doing so makes sense: that cable is a valuable conduit and if it can serve another purpose at the same time—that seems like a good idea. It’s a technique that has been used for many years, for example, to provide DC power to a microphone with an internal preamp via the existing audio cable, and also to provide DC power to the LNB (low noise block) preamp mounted at the feed of a consumer satellite dish.

It is not just the analog signal cables of a microphone or satellite front end that can also carry power. USB started out as a power path of modest capacity (about 10W) while Ethernet was never really a power path. Now, both USB via USB C and Type 3 standards, and power over Ethernet (PoE) via the almost-ready IEEE 802.3bt (often called PoE++) standard, raise the power level to about 100W while also supporting MBps and Gbps data rates. Note that PoE is not just a dream; it has been used successfully for area lighting in a school, as explained in this January 2017 article from Cabling Installation and Management, "Intelligent building cabling solutions support PoE lighting for a better education environment."

Such dual functionality is not restricted to adding the power onto a data cable, but can be used in the complementary mode. There are several standards, both formal and propriety, for using the ubiquitous AC line as a data carrier. These work, but with slower speeds than many applications can accept, and significant issues related to blocked pathways due to transformers, horrendous system noise and transients, and more.

Supporting both power and data isn’t an easy task. To merge them onto that single tempting copper cable requires a lot of circuitry at both ends to set up the connection, establish a protocol, verify that all’s well, and watch for problems which may occur before or during the connection, including connecting while hot as well as literally pulling the plug. Further, the cable and connectors must meet some strignent electrical and mechanical standards. Finally, coupling both the data and power to the copper is a non-trivial exercise, in some cases requiring special inductive components.

So, is it worth the effort, cost, and design complexity? As with most engineering tradeoffs, the answer is “it depends.” It’s a function of user priorities in terms of ease of use, potential simplification, acceptable confusion, troubleshooting effort, cost, consistency of operation, and many other factors.

I am somewhat skeptical, at least, for several reasons. First, added complexity of the interconnect means more things to go wrong across active hardware, software, and the cable/connector assembly. Using these will not be a “no-brainer” since not just any USB cable or Ethernet connector will work; they have to meet the latest specifications.

The compromise and limitation factors also are an issue, as many times, a single-function, optimized solution is actually a better choice than one designed for multiple functions. It’s like having a basic screwdriver versus the Swiss Army knife which has many tools on a single handle. The latter is unquestionably handy especially in a crisis, but that single-function screwdriver is probably the preferred tool under normal circumstances: it does one thing, it does it well, and it has far fewer limitations. Or consider the more technical example of the often-hyped flying car, which seems to re-appear every few years as “the next big thing.” It is capable of both functions, but is not really good at either, and needs many design and functional compromises to make it viable.

It will be interesting to see where and for whom these data + power, single cable approaches work well, and if they cause more headaches than the problems they were intended to solve. I wonder if either one will become a dead-end as problems and frustrations outweigh the gains they bring? It will an interesting situation to watch as not every standard succeeds, even with major industry backing.

Is there a standard which you thought would be a winner, but it did not gain wide acceptance? What’s your view on the viability of PoE++ and USB Type 3 for power delivery versus a separate higher-power cable?

Also see:

Loading comments...

Write a Comment

To comment please Log In